SafeArray problem

pixie888 at hotmail.com pixie888 at hotmail.com
Thu Jun 20 09:11:01 CEST 2002


On Wed, 19 Jun 2002 11:28:36 -0400, "Michael Weiss"
<FooWeissBarMike at hotmail.com> wrote:

>
>> I have not written an idl-file myself, I just used the classwizard in
>> Visual Studio to write my method on the automation object (using MFC).
>> What do you really want to know? (I looked for idl-files in my project
>> but did not find any)
>
>How about an ODL file?

not found too, I guess that when you use MFC in combination with the
Classwizard things are handled hidden for the programmer. Nonetheless:
I give you the following information for the better of it:

im my .h file

afx_msg LPDISPATCH GetRelationQuerierEx(const VARIANT FAR& vaIDs,long
lDummy);

the first argument is the safearray, the second the long

in my .cpp file

DISP_PROPERTY_PARAM(CDispCore, "RelationQuerierEx",
GetRelationQuerierEx, SetNotSupported, VT_DISPATCH, VTS_VARIANT
VTS_I4)

>
>I was just wondering if the in, in/out, retval, pointer vs. non-pointer
>symantics of the parameter were wrong. I was guessing by adding a new
>parameter (the long) that the safearray's properties were changed for the
>better. But I guess with a pure Dispatch interface (dispinterface) that this
>kind of stuff isn't done, or at least its done differently.
>
>Still, I wonder if this still isn't the problem of the function not being
>declared correctly before you made the change and then when you added that
>long the declaration of the safearray was changed.
>
>
>
>
>




More information about the Python-list mailing list